Embryo quality comprehensive evaluation device based on deep learning

Abstract

The invention discloses an embryo quality comprehensive evaluation device based on deep learning. The device comprises a computer memory, a computer processor and a computer program which is stored inthe computer memory and can be executed on the computer processor, an embryo quality comprehensive evaluation model is stored in the computer memory, and when the computer processor executes the computer program, the following steps are realized: receiving an embryo image; identifying and segmenting a regional image of the cleavage ball through a cleavage ball target detection module; using the cell counting module to count cells in the regional image, and obtaining the development stage of the cleavage ball; using the embryo quality analysis module to score the development condition of the cleavage ball in the regional image; and in the prediction module, scoring the implantation possibility of the cleavage ball according to the regional image, and obtaining a prediction result of the implantation success rate of the cleavage ball by integrating the scores of the development conditions. The prediction result can assist a doctor in predicting the success rate of embryo implantation.

Description

technical field [0001] The invention relates to the field of artificial intelligence medical devices, in particular to a comprehensive evaluation device for embryo quality based on deep learning. Background technique [0002] Reproduction is one of the key elements in the continuation of life. In recent years, the incidence of reproductive disorders has increased year by year, which has a potential and far-reaching impact on the quality of life of men and women, and has become a problem that threatens human health. Since the world's first test-tube baby, Louis Brown, was born in the UK on July 25, 1978, assisted reproductive technology has been widely used around the world and has become an effective way to help humans deal with this problem. This technology uses in vitro fertilization-embryo transfer as a treatment for fallopian tube diseases, male factor, anovulation and unexplained infertility. The main steps include gamete preparation, in vitro fertilization, embryo cul...

AI Technical Summary

Problems solved by technology

[0004] 1. Lack of reliability and consistency

Due to individual differences, embryologists have a certain degree of subjectivity in judging the evaluation criteria, which leads to unavoidable reliability and lack of consistency in artificial scoring, which will interfere with the judgment of embryo quality.

A survey shows that even embryologists trained under the same standard have a high scoring consistency when the embryos are in good shape, but the scoring consistency will decrease when the embryos are in poor shape and have large variations

[0005] 2. Unquantifiable

[0006] 3. Inconsistent standards

[0007] 4. It is difficult to identify the number of blastomeres

However, the products of cell division and cell debris existing in the process of human embryo division cover the blastomere cells, which interferes with the identification of the number of cells

In the case of high number of blastomeres and high fragmentation, embryologists cannot identify the number of cleavage well

Images